Purification and analysis of expression of the stationary phase-inducible slp lipoprotein in Escherichia coli: role of the Mar system.

Entrance into the stationary phase in Escherichia coli induces a group of stress response genes including slp, which encodes an outer membrane lipoprotein. Glucose limitation is sufficient, but not necessary, for the increase in slp expression. The Slp protein was purified and an antibody-based assay was developed, which enabled identification of Slp as spot R2226 (G018.1) in the E. coli gene-protein database. Although Slp is a major membrane component in stationary phase cultures grown in complex medium, no significant changes in resistance to oxidative stress or membrane perturbants were found in a slp null mutant strain. The presence or absence of multiple antibiotic resistance (Mar) A did not alter the final stationary phase levels of Slp. However, the Mar system could modestly influence the level of slp expression during the transition from exponential to stationary growth phase. Accumulation of Slp leads to a small increase in sensitivity to chloramphenicol.

Characterization of the carbon starvation-inducible and stationary phase-inducible gene slp encoding an outer membrane lipoprotein in Escherichia coli.

Escherichia coli induces the expression of more than 50 proteins in response to starvation for a carbon source. Strains MC7 (csi7::phoA) and MC19 (csi19::phoA) contain fusions of a signal peptide-deficient phoA reporter sequence to a csi (carbon starvation-inducible) gene. PhoA expression increased when these strains were deprived of a carbon source or entered stationary phase but did not when the cells were deprived of a nitrogen source or subjected to osmotic, oxidative or thermal stress. Mapping and sequence analysis of the cloned phoA fusions in strains MC7 and MC19 indicated that they had occurred in different locations within the same previously unidentified gene. The wild-type allele of this gene was cloned and the encoded protein was found to be a new lipoprotein. Therefore we propose to call this locus slp (starvation lipoprotein). The 22 kDa Slp protein is associated with the outer membrane fraction. The slp gene was located at 78.6 centisomes on the E. coli genetic map. The -10 and -35 regions upstream of the mRNA start site were characteristic of a sigma 70 promoter. The major transcript from this promoter was sufficiently large to contain slp sequences but not the downstream open reading frame. Induction of beta-galactosidase activity from a slp::lacZ translational fusion during carbon starvation or stationary phase was independent of cAMP, RpoS (KatF) and DnaK, all of which are known to affect the expression of certain starvation-inducible or stationary phase-inducible proteins.

Role of Clp protease subunits in degradation of carbon starvation proteins in Escherichia coli.

When deprived of a carbon source, Escherichia coli induces the synthesis of a group of carbon starvation proteins. The degradation of proteins labeled during starvation was found to be an energy-dependent process which was inhibited by the addition of KCN and accelerated when cells were resupplied with a carbon source. The degradation of the starvation proteins did not require the ATP-dependent Lon protease or the energy-independent proteases protease I, protease IV, OmpT, and DegP. During starvation, mutants lacking either the ClpA or ClpP subunit of the ATP-dependent Clp protease showed a partial reduction in the degradation of starvation proteins. Strains lacking ClpP failed to increase degradation of starvation proteins when glucose was added to starving cells. The clpP mutants showed a competitive disadvantage compared with wild-type cells when exposed to repeated cycles of carbon starvation and growth. Surprisingly, the glucose-stimulated, ClpP-dependent degradation of starvation proteins did not require either the ClpA or ClpB protein. The patterns of synthesis of starvation proteins were similar in clpP+ and clpP cells. The clpP mutants had reduced rates of degradation of certain starvation proteins in the membrane fraction when a carbon source was resupplied to the starved cells.

Effects of production of abnormal proteins on the rate of killing of Escherichia coli by streptomycin.

The role of abnormal membrane proteins in modulating the rate of killing by streptomycin was investigated. Davis et al. (B.D. Davis, L. Chen, and P.T. Tai, Proc. Natl. Acad. Sci. USA 83:6164-6168, 1986) have proposed that misread membrane proteins created by the action of streptomycin on translating ribosomes cause the formation of nonspecific membrane channels which allow increased uptake of the antibiotic and contribute to its bactericidal action. Pretreatment of Escherichia coli with a low concentration of puromycin enhanced the rate of killing by streptomycin. The effect of the pretreatment with puromycin was transient, since approximately normal rates of killing by streptomycin were restored after 30 min of incubation in antibiotic-free medium. This time period correlates with the time required to degrade labile polypeptides in puromycin-treated cells. The induction of a specific abnormal malE-lacZ fusion protein, which is capable of disrupting the normal membrane protein secretion process, also increased the rate of killing by streptomycin. Induction of malF-phoA fusion proteins, which have no significant effects on membrane integrity, did not alter susceptibility to streptomycin. These observations suggest that certain abnormal membrane proteins can contribute to the bactericidal action of streptomycin.

Characterization of a membrane-associated serine protease in Escherichia coli.

Three membrane-associated proteolytic activities in Escherichia coli were resolved by DEAE-cellulose chromatography from detergent extracts of the total envelope fraction. On the basis of substrate specificity for the hydrolysis of chromogenic amino acid ester substrates, the first two eluting activities were determined previously to be protease V and protease IV, respectively (M. Pacaud, J. Bacteriol. 149:6-14, 1982). The third proteolytic activity eluting from the DEAE-cellulose column was further purified by affinity chromatography on benzamidine-Sepharose 6B. We termed this enzyme protease VI. Protease VI did not hydrolyze any of the chromogenic substrates used in the detection of protease IV and protease V. However, all three enzymes generated acid-soluble fragments from a mixture of E. coli membrane proteins which were biosynthetically labeled with radioactive amino acids. The activity of protease VI was sensitive to serine protease inhibitors. Using [3H]diisopropylfluorophosphate as an active-site labeling reagent, we determined that protease VI has an apparent molecular weight of 43,000 in polyacrylamide gels. All three membrane-associated serine proteases were insensitive to inhibition by Ecotin, and endogenous, periplasmic inhibitor of trypsin.

Evidence for RecA protein association with the cell membrane and for changes in the levels of major outer membrane proteins in SOS-induced Escherichia coli cells.

Membrane fractions from Escherichia coli cells expressing DNA damage-inducible (SOS) functions contain elevated quantities of RecA protein (L. J. Gudas and A. B. Pardee, J. Mol. Biol. 101:459-477, 1976). We used two-dimensional polyacrylamide gel electrophoresis to separate membrane proteins from several strains to determine whether this effect is an artifact due to contamination of membranes during preparation by the large amount of cytoplasmic RecA present in SOS-induced cells. We found that amplification of RecA+ protein without a DNA-damaging treatment does not result in increased RecA-membrane association, whether recA is depressed specifically by an operator-constitutive recA allele or coordinately with other SOS genes by a lexA mutation that inactivates their common repressor. In contrast, large amounts of RecA appear in membrane fractions from undamaged cells of an SOS-constitutive strain carrying recA730, which encodes a spontaneously SOS-activated RecA. We conclude that the increased association of RecA with the membrane fraction requires the presence of the activated form of RecA, and that this association may contribute significantly to the SOS response. We describe also striking effects of SOS expression on the levels of the outer membrane proteins OmpA, OmpC, and OmpF.

Relative stability of membrane proteins in Escherichia coli.

The relative stability of membrane proteins in Escherichia coli was investigated to determine whether these proteins are degraded at heterogeneous rates and, if so, whether the degradative rates are correlated with the sizes or charges of the proteins. Cells growing in a glucose-limited chemostat with a generation time of 15 h were labeled with [(14)C]leucine. After allowing 24 h for turnover of (14)C-labeled proteins, the cells were labeled for 15 min with [(3)H]leucine. By this protocol, the rapidly degraded proteins have a high ratio of (3)H to (14)C, whereas the stable proteins have a lower ratio. The total cell envelope fraction was collected by differential centrifugation, and the proteins were separated by two-dimensional polyacrylamide gel electrophoresis. The relative ratio for each protein was determined by dividing its (3)H/(14)C ratio by the (3)H/(14)C ratio of the total membrane fraction. Although most of the 125 membrane proteins had relative ratios close to the average for the total membrane fraction, 19 varied significantly from this value. These differences were also observed when the order of addition of [(14)C]leucine and [(3)H]leucine was reversed. In control cultures labeled simultaneously with both isotopes, the relative ratios of these 19 proteins were similar to that of the total membrane fraction. Thirteen of these proteins had low relative ratios, which suggested that they were more stable than the average protein. An experiment in which the normal labeling procedure was followed by a 60-min chase period in the presence of excess unlabeled leucine suggested that the low relative ratios of 3 of these 13 proteins may be due to a slow post-translational modification step. Six membrane proteins had high relative ratios, which indicated that they were degraded rapidly. In contrast to the relationships found for soluble proteins in mammalian cells, there were no strong correlations between the degradative rates and either the isoelectric points or the molecular weights of membrane proteins in E. coli.

Properties of the lysosomal apparatus during differentiation of cultured striated muscle cells.

A technique for the preparation of relatively pure myoblasts from chick primary culture is described. Cultured rat myogenic cells (L6) were plated and grown to provide three morphologically distinct cell populations: perfusion, postfusion, and nonfusion. Homogenates of L6 cells demonstrated two major peaks of proteolytic activity at pH 3.0 and 5.5. The activity could be partially inhibited by leupeptin or pepstatin. Differential centrifugation indicated significant acid hydrolase activity in the "H" fraction of prefused cells, which shifted to the "L" fraction after fusion of the cells. Two populations of lysosomes were resolved in the myoblasts and myotubes after isopycnic centrifugation in Percoll. The equilibrium densities were 1.044 and 1.060-1.068. Cells were incubated with several protease inhibitors. Only chloroquine caused a large inhibition of protein degradation.

Relative stability of intracellular proteins in bacterial cells.

The relative stabilities of soluble and membrane proteins were examined in growing Escherichia coli cells. In contrast to mammalian cells, we found no correlations between the isoelectric points or molecular weights of E. coli proteins and their degradative rates. The soluble proteins with short half-lives tended to be degraded preferentially in vitro by trypsin or chymotrypsin. The stability of membrane proteins in vivo was correlated with in vitro sensitivity to chymotrypsin but not to trypsin. In the total membrane fraction, endogenous proteolytic activity varied with growth conditions. This activity was inhibited by o-phenanthroline, EDTA and dithiothreitol suggesting that one or more metallo-proteinases were present. Membrane proteinase activity was also inhibited by phenethyl alcohol, a membrane perturbant. The abundance of the membrane proteins that were most labile in vivo was dependent on growth conditions. The most labile protein accumulated in the outer membrane with an inverse relationship to growth rate.

Effects of starvation for potassium and other inorganic ions on protein degradation and ribonucleic acid synthesis in Escherichia coli.

Starvation of Escherichia coli for potassium, phosphate, or magnesium ions leads to a reversible increase in the rate of protein degradation and an inhibition of ribonucleic acid (RNA) synthesis. In cells deprived of potassium, the breakdown of the more stable cell proteins increased two- to threefold, whereas the hydrolysis of short-lived proteins, both normal ones and analog-containing polypeptides, did not change. The mechanisms initiating the enhancement of proteolysis during starvation for these ions were examined. Upon starvation for amino acids or amino acyl-transfer RNA (tRNA), protein breakdown increases in relA+ (but not relA) cells as a result of the rapid synthesis of guanosine-5'-diphosphate-3'-diphosphate (ppGpp). However, a lack of amino acyl-tRNA does not appear to be responsible for the increased protein breakdown in cells starved for inorganic ions, since protein breakdown increased in the absence of these ions in both relA+ and relA cultures, and since a large excess of amino acids did not affect this response. In bacteria in which energy production is restricted, ppGpp levels also rise, and protein breakdown increases. The ion-deprived cultures did show a 40 to 75% reduction in adenosine-5'-triphosphate levels,l similar to that seen upon glucose starvation. However, this decrease in ATP content does not appear to cause the increase in protein breakdown or lead to an accumulation of ppGpp. No consistent change in intracellular ppGpp levels was found in relA+ or relA cells starved for these ions. In addition, in relX mutants, removal of these ions led to accelerated protein degradation even though relX cells are unable to increase ppGpp levels or proteolysis when deprived of a carbon source. In the potassium-, phosphate-, and magnesium-deprived cultures, the addition of choramphenicol or tetracycline caused a reduction in protein breakdown toward basal levels. Such findings, however, do not indicate that protein synthesis is essential for the enhancement of protein degradation, since blockage of protein synthesis by inactivation of a temperature-sensitive valyl-tRNA synthetase did not restore protein catabolism to basal levels. These various results and related studies suggest that the mechanism for increased protein catabolism on starvation for inorganic ions differs from that occurring upon amino acid or arbon deprivation and probably involves an enhanced susceptibility of various cell proteins (especially ribosomal proteins) to proteolysis.

Degradation of proteins in steady-state cultures of Escherichia coli.

The degradation of proteins in Escherichia coli was investigated in cells grown under steady-state conditions in a glucose-limited chemostat. During the first 24 h, approximately 25% of pulse-labeled proteins were degraded and after 72 h up to 58% of the proteins were broken down. To examine the stability of subcellular components steady-state cultures were labeled with an initial pulse of [14C]leucine, 24 h were allowed for turnover of these proteins, and the cells were then labeled with a short pulse of [3H]leucine. By this double-label protocol, the labile proteins were preferentially labeled with [H]leucine and had high 3H/14C ratios, while the more stable proteins had lower 3//14C ratios. The 3/-labeled proteins were degraded approximately five times as rapidly as the 14C-labeled proteins in exponentially growing cells. The relative stability of subcellular fractions was determined by comparing their 3H/14C ratios to the ratio of the cells at harvest. The soluble fraction contained the most labile proteins, while the ribosomal and membrane fractions were at least as stable as the average cell protein.

Bacteriophages inhibit degradation of abnormal proteins in E. coli.

On infection of Escherichia coli cells by bacteriophages T4, T5 or T7, the degradation of E. coli protein fragments and abnormal proteins is inhibited. Normal E. coli proteins, however, continue to be degraded at their usual rates. T4 early proteins (s) is needed to inhibit the turnover of abnormal proteins in T4-infected E. coli cells.

Macromolecular synthesis and cell division during morphogenesis of Arthrobacter crystallopoietes.

The sphere-rod-sphere morphology cycle of Arthrobacter crystallopoietes was accompanied by changes in the rate of growth and the rates of DNA, RNA and protein synthesis. The patterns of macromolecule synthesis resembled those found in other bacteria during a step-up followed by a step-down in growth rate. During the step-up in growth spherical cells grew into rods and macromolecules were synthesized in the absence of cell division. During step-down, successive rounds of septation produced progressively smaller cells which did not separate and remained in chains. The morphology of the cells was dependent on the growth rate and could be altered by changing the dilution rate in the malate-limited chemostat. Gradual transitions in morphology and gradual increases in macromolecule content of the cells occurred as the growth rate was increased in the chemostat. Sphere to rod morphogenesis occurred when DNA synthesis was inhibited by treatment with mitomycin C or by thymine starvation. The DNA-deficient rods did not divide and eventually lysed. DNA, RNA and protein synthesis were continuously required for the reductive division of rods to spheres.